Choose timezone
Your profile timezone:
120th Impedance Working Group Meeting (PML in Wakis)
120th Impedance Working Group Meeting (PML in Wakis)
Date: 23/09/2026
Present: C. Antuono, P. Bampis, R. Calaga, S. Calatroni, E. De La Fuente Garcia, A. Huber, A. Kyriakidou, P. Krkotic, E. Macchia, I. Mases Sole, N. Mounet, M. Neroni, H. Pommerenke, J. Sonpar, C. Zannini
Improvement of Wakis absorbing boundary condition (A. Huber)
Carlo comments on the challenges related to the implementation of the PML in Wakis and congratulates Antonia on her work.
Nicolas comments on the implementation of the CPML as a dielectric constant (kappa) + relaxation time, saying that it actually looks physical. Elena DLFG comments that the explanation was simplified for this presentation, but kappa is affecting the derivatives, so it's not like a dielectric constant: looking at how it is introduced in the time-domain equations, there is a convolution stretching the derivatives of the fields. Nicolas observes it is anisotropic and Elena DLFG confirms. *Antonia* adds that it affects mu and epsilon at the same time.
Nicolas asks if imposing eta=1 to have perfectly matched layer is a Leontovich-like approximation or if is it correct; Antonia answers that theoretically it is correct, it's just numerically that it is not continuous.
Patrick asks if the user needs to understand how many layers are needed, Antonia comments that it is implemented in Wakis so the user does not need to touch it (and typically 4 layers should suffice); but if there are a lot of evanescent fields, then the user should make considerations the parameters.
Patrick asks about injection schemes, Antonia comments that the implemented injection scheme should be similar to the analytical injection scheme in CST, but we can only guess since CST is like a black box to its users. Antonia adds that her implementation only works for beta=1 while probably CST has a more advanced implementation; Carlo comments that in CST for beta<1 users are forced to use the analytical injection scheme (and cannot use the transmission line injection scheme).
Nicolas asks if everything is based on the assumption that above cut-off everything is absorbed, or if only a few discrete frequencies (which can really propagate in the waveguide) are absorbed. Carlo and Chiara observe that all the modes, above their cut-off frequency are propagating modes for every higher frequency.
Carlo comments that Wakis is now ready to go for lossless simulations (geometric impedance). Elena DLFG adds that everything is fully merged and available on GitHub.
AoBs
Michela says that there is no new ECR to be commented by IWG.
There was the ECR on the reduction of heat load in the arcs, but it was already discussed by the IWG; Nicolas comments that he checked and we are on version 0.2 and he already commented on it.
Regarding the ECR on BFA09, simulations were performed by Hermann and its comment was implemented in the new version of the ECR.
Carlo informs that tomorrow there is going to be a review organized by BI for BGI and BSRT, to put everything together in general from the equipment side; he will give a presentation on impedance considerations.
There are minutes attached to this event.
Show them.